On the heels of this comment from JH: https://issues.apache.org/jira/browse/FLUME-735?focusedCommentId=13083189&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13083189
I'm trying to come up with the config pattern for disk failover for an hbase sink. Assuming my hbase sink was called 'myhbase' would I do this? < myhbase ? diskFailover myhbase > or would I need to include all the other insistent/stubborn pieces like this: < myhbase ? diskFailover insistentAppend stubbornAppend insistentOpen myhbase > I was looking at the source but was getting lost in what this chain was actually doing. Hopefully Jon can explain his original thinking. Basically, I'm looking for local disk failover and retry for hbase outage/maintenance windows This is cdh3u3 flume (not NG). Thanks, Steve
